ai

Chat lifecycle middleware hooks: onConfig, onStart, onChunk, onBeforeToolCall, onAfterToolCall, onUsage, onFinish, onAbort, onError. Use for analytics, event firing, tool caching (toolCacheMiddleware), logging, and tracing. Middleware array in chat() config, left-to-right execution order. NOT onEnd/onFinish callbacks on chat() — use middleware.

Entry point for TanStack AI skills. Routes to chat-experience, tool-calling, media-generation, structured-outputs, adapter-configuration, ag-ui-protocol, middleware, custom-backend-integration, and debug-logging. Use chat() not streamText(), openaiText() not createOpenAI(), toServerSentEventsResponse() not manual SSE, middleware hooks not onEnd callbacks.

Isomorphic tool system: toolDefinition() with Zod schemas, .server() and .client() implementations, passing tools to both chat() on server and useChat/clientTools on client, tool approval flows with needsApproval and addToolApprovalResponse(), lazy tool discovery with lazy:true, rendering ToolCallPart and ToolResultPart in UI.

Image, audio, video, speech (TTS), and transcription generation using activity-specific adapters: generateImage() with openaiImage/geminiImage, generateAudio() with geminiAudio/falAudio, generateVideo() with async polling (openaiVideo/geminiVideo/grokVideo/falVideo, per-model typed durations), generateSpeech() with openaiSpeech, generateTranscription() with openaiTranscription. React hooks: useGenerateImage, useGenerateAudio, useGenerateSpeech, useTranscription, useGenerateVideo. TanStack Start server function integration with toServerSentEventsResponse.

Server-side AG-UI streaming protocol implementation: StreamChunk event types (RUN_STARTED, TEXT_MESSAGE_START/CONTENT/END, TOOL_CALL_START/ARGS/END, RUN_FINISHED, RUN_ERROR, STEP_STARTED/STEP_FINISHED, STATE_SNAPSHOT/DELTA, CUSTOM), toServerSentEventsStream() for SSE format, toHttpStream() for NDJSON format. For backends serving AG-UI events without client packages.

Provider adapter selection and configuration: openaiText, anthropicText, geminiText, ollamaText, grokText, groqText, openRouterText, bedrockText, openaiCompatible. Per-model type safety with modelOptions, reasoning/thinking configuration, runtime adapter switching, extendAdapter() for custom models, createModel(). Generic OpenAI-compatible providers (DeepSeek, Together, Fireworks, etc.) via openaiCompatible({ baseURL, apiKey, models }) from @tanstack/ai-openai/compatible. API key env vars: OPENAI_API_KEY, ANTHROPIC_API_KEY, GOOGLE_API_KEY/GEMINI_API_KEY, XAI_API_KEY, GROQ_API_KEY, OPENROUTER_API_KEY, OLLAMA_HOST, BEDROCK_API_KEY (or AWS_BEARER_TOKEN_BEDROCK).

Run harness adapters (Claude Code, Codex, OpenCode) INSIDE isolated sandboxes via defineSandbox + withSandbox + a provider (localProcessSandbox / dockerSandbox). Covers declarative provisioning: createSecrets + secret/bearer, skills (agentSkill/gitSkill/mcpSkill/ fileSkill), plugins, instructions → canonical AGENTS.md + symlinks projected per harness; shallow-clone default with depth opt-out; serial/parallel setup callback over a persistent shell; snapshot-after-setup default with snapshotMaxAge TTL; defineWorkspace (git/setup/scripts/skills/secrets/ instructions/plugins), defineSandboxPolicy (allow/ask/deny), lifecycle/resume, the SandboxHandle (fs/git/process/ports), capability tokens, defineSandbox hooks (onFile/onFileCreate/onFileChange/onFileDelete/onReady/onError/ onDestroy) + fileEvents flag, chat middleware sandbox group (defineChatMiddleware sandbox hooks), the sandbox debug category, watchWorkspace as a low-level building block, and the file.changed / sandbox.file / claude-code.session-id events. Us

Audit TanStack AI provider adapters for feature parity gaps and outdated model lists. Triggered as /gap-analysis <provider|feature <name>|models|--all>. Produces a dated markdown report under .agent/gap-analysis/. Maintainer tool — does not edit feature-support.ts or model-meta.ts directly.

Host-side Model Context Protocol (MCP) client for TanStack AI: connect to external MCP servers, discover and run their tools inside any adapter's chat() loop, read resources and prompts, generate TypeScript types (typed tool names/pool keys) with the bundled CLI, and manage lifecycle with close()/await using.

End-to-end chat implementation: server endpoint with chat() and toServerSentEventsResponse(), client-side useChat hook with fetchServerSentEvents(), message rendering with UIMessage parts, multimodal content, thinking/reasoning display. Covers streaming states, connection adapters, and message format conversions. NOT Vercel AI SDK — uses chat() not streamText().

LLM-generated TypeScript execution in sandboxed environments: createCodeModeTool() with isolate drivers (createNodeIsolateDriver, createQuickJSIsolateDriver, createCloudflareIsolateDriver), codeModeWithSkills() for persistent skill libraries, trust strategies, skill storage (FileSystem, LocalStorage, InMemory, Mongo), client-side execution progress via code_mode:* custom events in useChat.

Connect useChat to a non-TanStack-AI backend through custom connection adapters. ConnectConnectionAdapter (single async iterable) vs SubscribeConnectionAdapter (separate subscribe/send). Customize fetchServerSentEvents() and fetchHttpStream() with auth headers, custom URLs, and request options. Import from framework package, not @tanstack/ai-client.

Pluggable, category-toggleable debug logging for TanStack AI activities. Toggle with `debug: true | false | DebugConfig` on chat(), summarize(), generateImage(), generateSpeech(), generateTranscription(), generateVideo(). Categories: request, provider, output, middleware, tools, agentLoop, config, errors. Pipe into pino/winston/etc via `debug: { logger }`. Errors log by default even when `debug` is omitted; silence with `debug: false`.

Type-safe JSON schema responses from LLMs using outputSchema on chat() and useChat(). Supports Zod, ArkType, and Valibot schemas. The adapter handles provider-specific strategies transparently — never configure structured output at the provider level. Pass stream:true alongside outputSchema for incremental JSON deltas + a terminal validated object via the `structured-output.complete` event. Every assistant turn in useChat carries its own typed `StructuredOutputPart` on `messages[i].parts`, so multi-turn structured chats preserve history automatically — partial/final derive from the latest assistant turn's part. convertSchemaToJsonSchema() for manual schema conversion.

Triage all open GitHub issues, PRs, and discussions in the current repository by fanning out up to 100 parallel subagents (one per item), then produce a single prioritized report ranking which PRs to review first, which issues to address first, and which discussions need maintainer attention. Use when the user asks to "triage open issues/PRs", "triage discussions", "prioritize the backlog", "what should I review first", "sweep the repo", or any request to bulk-evaluate open GitHub work and recommend an order.